Browsing by Author "Strahl, Jonathan"

Now showing 1 - 6 of 6

Directing and Combining Multiple Queries for Exploratory Search by Visual Interactive Intent Modeling
(2021) Strahl, Jonathan; Peltonen, Jaakko; Floreen, Patrik
Conference article in proceedings
In interactive information-seeking, a user often performs many interrelated queries and interactions covering multiple aspects of a broad topic of interest. Especially in difficult information-seeking tasks the user may need to find what is in common among such multiple aspects. Therefore, the user may need to compare and combine results across queries. While methods to combine queries or rankings have been proposed, little attention has been paid to interactive support for combining multiple queries in exploratory search. We introduce an interactive information retrieval system for exploratory search with multiple simultaneous search queries that can be combined. The user is able to direct search in the multiple queries, and combine queries by two operations: intersection and difference, which reveal what is relevant to the user intent of two queries, and what is relevant to one but not the other. Search is directed by relevance feedback on visualized user intent models of each query. Operations on queries act directly on the intent models inferring a combined user intent model. Each combination yields a new result (ranking) and acts as a new search that can be interactively directed and further combined. User experiments on difficult information-seeking tasks show that our novel system with query operations yields more relevant top-ranked documents in a shorter time than a baseline multiple-query system.
Matrix factorization model for predicting user preference
(2020-06-23) Tähkä, Iiro
Perustieteiden korkeakoulu | Bachelor's thesis
Matrix Factorization With Graph Side Information
(2019-04-24) Laine, Eetu
Perustieteiden korkeakoulu | Bachelor's thesis
Negative relevance feedback for exploratory search with visual interactive intent modeling
(2017-03-07) Peltonen, Jaakko; Strahl, Jonathan; Floréen, Patrik
A4 Artikkeli konferenssijulkaisussa
In difficult information seeking tasks, the majority of topranked documents for an initial query may be non-relevant, and negative relevance feedback may then help find relevant documents. Traditional negative relevance feedback has been studied on document results; we introduce a system and interface for negative feedback in a novel exploratory search setting, where continuous-valued feedback is directly given to keyword features of an inferred probabilistic user intent model. The introduced system allows both positive and negative feedback directly on an interactive visual interface, by letting the user manipulate keywords on an optimized visualization of modeled user intent. Feedback on the interactive intent model lets the user direct the search: Relevance of keywords is estimated from feedback by Bayesian inference, influence of feedback is increased by a novel propagation step, documents are retrieved by likelihoods of relevant versus non-relevant intents, and the most relevant keywords (having the highest upper confidence bounds of relevance) and the most non-relevant ones (having the smallest lower confidence bounds of relevance) are shown as options for further feedback. We carry out task-based information seeking experiments with real users on difficult real tasks; we compare the system to the nearest state of the art baseline allowing positive feedback only, and show negative feedback significantly improves the quality of retrieved information and user satisfaction for difficult tasks.
Patient appointment scheduling system: with supervised learning prediction
(2015-06-10) Strahl, Jonathan
Perustieteiden korkeakoulu | Master's thesis
Large waiting times at hospital outpatient clinics are a cause of dissatisfaction to patients, cause additional stress to hospital staff, increase the risk of contagion and add complications for patients with medical conditions. Reducing waiting times and surgeon idle time improves the quality of service and efficiency of a hospital: this is a recently growing focus in healthcare. Oulu hospital wants to identify and reduce large waiting times at their out-patient clinic. For the past few years the clinic has used a self-service system whereby patients register on arrival and hospital staff use a patient call-in system. The past schedules are analyzed using this data: information visualizations and performance measures are provided. The worst performing clinic sessions are the subject of the scheduling optimization prototype system. The scheduling optimization focuses on predicting the duration of an appointment and the late arrival of the surgeon. These two factors have been identified as causes of long patient waiting times. The variance of the duration is identified to be high, therefore supervised-learning regression is used for both simple inference and prediction. The features that are good predictors and the results of the prediction accuracy are reported. With the predicted appointment durations, and surgeon arrival times, a scheduling optimization approach is used to improve the existing schedule; a simple greedy hill-climbing approach is evaluated. It is found that using the historical data to simulate a real day, appointment rules and scheduling optimization the patient waiting time is reduced with this method. Showing the system to be potentially promising.
Predicting treatment durations for improving health care schedules
(2019-12-16) Forsman, Anders
Perustieteiden korkeakoulu | Master's thesis
Utilization of healthcare resources is an important issue that has been extensively studied. Healthcare is expensive and as it is often funded by the government the allocation of resources is a contentious issue. This is pushing for maximizing the efficiency of the available resources. But, in addition, the patient experience is also important, and too long waiting times and queues can have a detrimental effect on patients and is generally not an enjoyable experience, especially when sick. In addition, waiting rooms for holding waiting patients is also an added cost. Much of the research has been aimed at the scheduling problem of how to match patients with professionals. However, in order to be able to efficiently making a working schedule, information on the actual treatment durations is valuable. Using machine learning to predict treatment durations is a research area that is not greatly covered in literature. And thus, this work focuses on analyzing factors impacting the treatment durations, as well as predicting the durations based on basic demographic and historic information about the patients. However, with the limited amount of relevant data available in this work, this is a great challenge and appears not to be easily achievable. In addition, an application for analyzing and exporting event data from a patient flow management system is created. The purpose of this application is to aid in debugging the main software, and make it easier to export different data set and conduct additional studies in the future.